A framework to assess national level vulnerability from theperspective of food security: The case of coral reef fisheries

Sara Hughes a,1,*, Annie Yau a,2, Lisa Max b, Nada Petrovic c,3, Frank Davenport d,Michael Marshall d,4, Timothy R. McClanahan e, Edward H. Allison f,5, Joshua E. Cinner g

aBren School of Environmental Science and Management, University of California, Santa Barbara, CA 93106, USAbDepartment of Ecology, Evolution and Marine Biology, University of California, Santa Barbara, CA 93106, USAcDepartment of Physics, University of California, Santa Barbara, CA 93106, USAdDepartment of Geography, University of California, Santa Barbara, CA 93106, USAeWildlife Conservation Society, Marine Programs, Bronx, NY 10460, USAfWorldFish Center, P.O. Box 500, GPO, 10670 Penang, MalaysiagARC Centre of Excellence for Coral Reef Studies, James Cook University, Townsville, QLD 4811, Australia

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8

a r t i c l e i n f o

Published on line 21 August 2012

Keywords:

Adaptive capacity

Exposure

Resilience

Sensitivity

Developing countries

Social-ecological systems

Climate change

a b s t r a c t

Measuring the vulnerability of human populations to environmental change is increasingly

being used to develop appropriate adaptation policies and management plans for different

economic sectors. We developed a national-level vulnerability index that is specific to food

security policies by measuring nations’ relative vulnerabilities to a decline in their coral reef

fisheries. Coral reef fisheries are expected to decline with climate and anthropogenic

disturbances, which may have significant consequences for food security. The vulnerability

measure was composed of exposure, sensitivity, and adaptive capacity indicators specific to

fisheries, reef management, and food security. The vulnerability index was used to evaluate

27 countries, as data required to fully populate the theoretical framework was limited. Of

these, Indonesia and Liberia were identified as most and Malaysia and Sri Lanka as least

vulnerable nations. Our analysis revealed two common national vulnerability characteriza-

tions: low income countries with low adaptive capacity and middle-income countries with

higher adaptive capacity but high sensitivity. These results suggest developing context-

specific policies and actions to build adaptive capacity in the low-income countries, and to

decrease sensitivity in middle-income countries. Comparing our food security evaluation to

a more general vulnerability approach shows that they produce different priority countries

and associated policies.

Published by Elsevier Ltd.

* Corresponding author. Tel.: +1 810 835 1748/303 497 2872; fax: +1 303 497 8401.E-mail address: shughes@ucar.edu (S. Hughes).

1 Present address: National Center for Atmospheric Research, FL2 3106D, Boulder, CO 80307, USA.2 Present address: Office of the Assistant Administrator, Office of Atmospheric Research, National Oceanic and Atmospheric Adminis-

tration, Silver Spring, MD 20910, USA.3 Present address: Center for Research on Environmental Decisions, 406 Schermerhorn-MC5501, Columbia University, New York, NY

10027, USA.4 Present address: United States Geological Survey, Flagstaff, AZ, USA.5

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/envsci

Present address: School of International Development, University of East Anglia, Norwich NR4 7TJ, UK.1462-9011/$ – see front matter. Published by Elsevier Ltd.http://dx.doi.org/10.1016/j.envsci.2012.07.012

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 896

1. Introduction

Enhancing food security is a rising concern and a central

aim for development agencies, such as the United States

Agency for International Development (USAID) and the

United Nations’ World Food Programme (WFP). Food

security is a concept with multiple definitions and possibili-

ties for interpretation (Dilley and Boudreau, 2001; Gregory

et al., 2005). The United Nations Food and Agriculture

Organization (FAO) definition of food security: ‘‘when all

people, at all times, have physical, social, and economic

access to sufficient, safe, and nutritious food that meets

their dietary needs and food preferences for an active and

healthy life’’ (FAO, 2011) is adopted in this paper. Improved

assessments of environmental and food production

changes, as well as the capacity of people to adapt, would

enable decision makers and governments to proactively

design and implement policies and aid strategies. While

substantial attention has been focused on the relationship

between food security, environmental change and terrestri-

al agriculture (e.g., Schmidhuber and Tubiello, 2007; Brown

and Funk, 2008; Luers et al., 2003), less research has

examined how fisheries declines will impact the food

security of countries dependent on those fisheries (Smith

et al., 2010). This is particularly so for the small island and

small coastal states of developing countries, where fisheries

associated with coral reefs are especially important (Burke

et al., 2011).

Coral reefs are undergoing large-scale ecological change

associated with climate change, pollution, and increasing

fishing effort (Pandolfi et al., 2011; McClanahan, 2002).

Because these fisheries are sources of protein, micronu-

trients and income, they provide a useful system for

evaluating food security and vulnerability to environmental

change (Hicks et al., 2009; Whittingham et al., 2003; Cinner

et al., 2012). On many poorly-managed coral reefs, the

yields, diversity, and catch per unit effort of associated

fisheries have been, and are predicted to continue declining

due to anthropogenic forces such as land use change,

pollution, ocean acidification, rising seawater temperatures,

and over-fishing (Smith and Buddemeier, 1992; Wilkinson

and Buddemeier, 1994; Newton et al., 2007; Halpern et al.,

2008). Consequently, the aim of our study is to develop a

framework to identify the most vulnerable regions, the

mechanisms creating this vulnerability, and the potential

policy interventions that may reduce this food security

vulnerability. We calculate vulnerability as the degree to

which a country is susceptible to a decline in coral reef

fisheries as a food source and its ability to respond to the

decline.

The theoretical and empirical literature on vulnerability of

coral reef fisheries and their contribution to food production is

reviewed and used to derive adaptive capacity, exposure, and

sensitivity metrics. Results from our food-security specific

framework are compared to those from the more generic Reefs

at Risk report (Burke et al., 2011). The specific features of

vulnerability and relevant policy considerations are evaluated

to consider appropriate methods for constructing vulnerabili-

ty metrics.

2. Conceptualizing vulnerability and foodsecurity

2.1. National scale vulnerability to environmental change

Vulnerability includes the ‘‘exposure and sensitivity of a

system to single or multiple stressors’’ (Smit and Wandel,

2006) and the capacity of that system to successfully adjust to

or capitalize on the effects of those stressors. The attention

devoted to conceptualizing and measuring human vulnerabil-

ity to environmental change has increased as human

population growth and climate change issues become

increasingly relevant and acute (Adger, 2006; Adger and Kelly,

1999; Handmer et al., 1999). Research has attempted to

generate either a universal framework for assessing vulnera-

bility or to measure vulnerability specific to particular types of

change, such that context-specific policies and management

can be developed (McClanahan et al., 2008, 2009). These efforts

have improved understanding of the contextual drivers and

dynamics of vulnerability and highlighted weaknesses and

potential tradeoffs between generic and highly specific

vulnerability indices.

Vulnerability assessments can be broad or specific. Broad

vulnerability assessments focus on multiple sectors or

globally defined policy areas but the implications for policy

interventions are often not focused enough to determine

implementation needs. Specific vulnerability assessments

target identified problems in order to recommend the specific

intervention and scale of policies needed to reduce vulnera-

bility (Leurs, 2005; Ionescu et al., 2009). For example, broad

assessments of threats to coral reefs, such as the World

Resource Institute’s Reefs at Risk project (Burke et al., 2011),

can lack the specific recommendations needed by policy

makers or governments to understand and manage environ-

mental change as it relates to immediate socio-economic

problems, such as food security.

The components of vulnerability to a given hazard –

exposure, sensitivity and adaptive capacity – vary across space

and time, regardless of the spatio-temporal scale of the

stressor (Turner et al., 2003). As a result, most vulnerability

analyses tend to be specific to a place and context while linked

across scales (Turner et al., 2003), and are most frequently

assessed and influential at national and global levels (Pelling

and Uitto, 2001; Brooks et al., 2005; Allison et al., 2009).

Globalization, trade and linked economies result in responses

and adaptive pathways to coastal environmental stressors

that exhibit high levels of connectivity and lend themselves to

analyses at the regional or national scales (Adger, 2006). The

costs of national level analyses can be large due to the scarcity

of comprehensive and high quality national data and the

difficulties of integrating the variability in vulnerability across

these large systems. However, the benefit of assessing

vulnerability at the national level is that the results can

influence national-level policy responses and adaptive man-

agement strategies. Allison et al. (2009) and Burke et al. (2011),

for example, conducted national-level assessments relevant

to coastal vulnerability. However, these studies did not

address sector-specific policy needs, which is a frequent

deficiency in the vulnerability literature. Bell et al. (2011)

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8 97

examined national level vulnerability in terms of food security

for Pacific Island countries and territories, but included all

fisheries and aquaculture and only examined vulnerability in

the face of one major driver (climate change). The aim of this

paper, therefore, is to develop a more specific framework and

methodology and provide an analysis of national level

vulnerability from a food security perspective, where coral

reef fisheries provide a useful case study.

2.2. Coral reefs, food security and environmental change

Coral reefs are tropical nearshore marine ecosystems that

are home to a high diversity of fish, invertebrates, algae, and

reef-building corals. Coral reef fisheries are defined here as

fisheries that harvest organisms associated with coral reefs

and exclude pelagic and other non-reef species even if they

are harvested nearshore. Coral reef fisheries are often

artisanal and subsistent and use low capital and low

technology to harvest both fish and invertebrates, primarily

for local consumption and secondarily for trade (Cinner and

McClanahan, 2006). While coral reefs are globally valued for

their high biodiversity, locally they are a productive and

easily accessible food resource for millions of people

(Kawarazuka and Bene, 2011). In Southeast Asia, for exam-

ple, coral reef fisheries generate US$ 2.4 billion (Burke et al.,

2002) while in the Caribbean they generate US$ 310 million

per year by providing a range of ecosystem services to the

economy and society (Burke and Maidens, 2004). As part of

national food systems around the world, coral reef fisheries

contribute to the food security of those countries. According

to Dulvy and Allison (2009), ‘‘catches by subsistence and

artisanal fisheries make up more than half of the essential

protein and mineral intake for over 400 million people in the

poorest countries in Africa and south Asia.’’ For example,

almost 60% of the animal protein of an average Indonesian

resident comes from fish (Dey et al., 2005) and in eight Pacific

Island countries and territories, 50–90% of animal protein in

the diet of rural communities comes from fish (Bell et al.,

2009).

Developing countries’ reliance on and use of subsistence

coral reef fisheries for food and income is complicated by

access arrangements and export issues. Subsistence fishers’

access to food and income can be limited by highly profitable,

large-scale fisheries with sufficient capital to compete globally

(Kent, 1997; Pauly et al., 2005; Atta-Mills et al., 2004). Declining

yields and increasingly restricted access may have serious

implications for food security in some parts of the world.

Policy and management efforts to address this issue will be

assisted by evaluating the national context as it relates to food

security.

3. Methods: measuring national levelvulnerability to coral reef fisheries decline

This paper scales national vulnerability to declining coral reef

resources to understand the causes of vulnerability. We used

data from a range of sources to quantify vulnerability as a

function of three components: a country’s exposure to

environmental change and disturbances, its sensitivity to this

change, and its adaptive capacity, or potential to respond to the

change (Adger, 2006).

3.1. Exposure

We define exposure as the degree to which a country’s coral

reef fisheries are subject to degradation due to anthropogenic

threats. Exposure was estimated using scores of cumulative

impact that quantify the threat to the world’s coral reefs from

a set of 38 categories. These categories encompass anthropo-

genic drivers of change in marine ecosystems and include

fishing, land use changes, invasive species, shipping, and

pollution (see Halpern et al. (2008) for detailed methods on

quantifying cumulative impact). Using a map of coral reefs

within each country’s exclusive economic zone (EEZ) pro-

duced by the World Conservation Monitoring Centre, we

calculated the cumulative impact scores for each 1-km2 grid of

a country’s coral reefs. These cumulative impact scores for

each grid were then averaged for a country’s coral reefs. A

higher cumulative impact score indicates a country whose

reefs are subject to greater exposure to anthropogenic threats.

3.2. Sensitivity

Sensitivity was defined as the degree to which a country is

dependent on coral reef fisheries for food. Our method for

quantifying sensitivity draws on a study by Luers et al. (2003)

that incorporates both the dependence on a food resource and

the proximity of the food system to some damage threshold.

Therefore, a country would be considered highly sensitive to a

change in coral reef fisheries if it is highly dependent on reef

fisheries for protein and also has inadequate protein sources.

We use the formula:

S ¼ Reef fisheries dependenceProtein intake relative to threshold

� �

¼ fjdW=dxjðW=W0Þ

� �

¼ Coral reef protein=Total proteinTotal protein=Protein threshold

where W is the well-being of the system indicated by the total

amount of daily protein a person consumes in grams, x is a

stressor or the decrease in the amount of coral reef protein

harvested (caused by the degradation of the reef), and W0 is a

threshold value of well-being below which the person is not

receiving enough daily protein. In this paper, we use the FAO

and WHO recommended minimum amount of daily protein

needed to maintain health (30 g protein per day) as our mea-

sure for W0. The change in W with respect to x, jdW/dxj,quantifies how much well-being, W, changes due to a pertur-

bation in the stressor. The ratio W/W0 quantifies the nearness

of the system to the threshold value—as this ratio declines the

system nears the threshold and the sensitivity increases.

We assumed a positive, linear relationship between coral

reef protein harvested and protein consumption, such that

jdW/dxj scales as the fraction of coral reef protein out of total

protein. We made this assumption because the functional

form of jdW/dxj is impossible to estimate with existing data: it

would require a mathematical relationship between the

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 898

decrease in amount of coral reef protein harvested and protein

consumption. However, this term still captures the core idea

that a country is more sensitive to reef degradation if a higher

percentage of its total protein is derived from reef fisheries.

Total protein consumption data from the FAOstat dataset

between 1995 and 2005 were used to calculate the sensitivity of

countries to declines in coral reef fisheries. Coral reef protein

was distinguished from total fish protein by excluding protein

from fish oil, pelagic fish, and two miscellaneous categories of

fish. Demersal fish, crustaceans, mollusks and cephalopods

are common to measures of total and reef protein. A higher

sensitivity value indicates a country is highly dependent on

coral reef protein as a protein source, and its protein

consumption is close to or below the minimum recommended

protein consumption level. Recent research (e.g., Kawarazuka

and Bene, 2010) highlights that the main direct nutritional

contribution of seafood to diets in all but the most fish-

dependent populations is through provision of micro-nutri-

ents (essential fatty acids, minerals, vitamins). However, data

on supply and demand for these is seldom available, and

protein consumption presents a reasonable proxy for nutri-

ent-dense foods.

3.3. Adaptive capacity

Adaptive capacity was defined as a country’s potential to

respond to changes in the contribution of reef fisheries to the

food system and ability to take advantage of or mitigate these

changes. We disaggregated adaptive capacity into four

categories: assets, flexibility, learning, and social organization

(Cinner et al., 2009).

3.3.1. AssetsAssets were defined as the resources a country has at its

disposal to assist the fisheries sector in responding to a decline

in reef fisheries. Assets are components of adaptive capacity

because they represent the ability of a country to leverage

resources in response to changes in the food system. We

measured assets as physical (man-made), financial and

natural. A country that is adapting well would have adequate

levels of all three types of assets in order to leverage resources

to adjust to change.

The percentage of a country’s population with access to

sanitation as reported by the World Bank in 2000 was used as a

measure of a country’s physical assets. This metric has been

shown to be a key indicator of vulnerability (Brooks et al., 2005)

as it represents the extent of man-made infrastructure that a

country’s residents have at their disposal and is a better proxy

than, for example, density of road networks, which does not

correlate as well with the development status of a country as it

is dependent on physical geography. As an indicator of a

country’s financial assets, we used average GDP per capita

from 1995 to 2005 as reported by the World Bank. As an

indicator of natural assets we used reef area per capita in each

country’s EEZ. We assumed that countries with a higher

percentage of the population with access to sanitation, higher

per-capita GDP, and a higher reef area per capita will have a

greater ability to cope with changes in coral reef fisheries via

compensatory mechanisms, alternative economic activities,

or new fishing methods.

3.3.2. FlexibilityFlexibility refers to the range of options a country has to meet

nutritional and livelihood needs if existing resource availabil-

ity declines. Flexibility is a component of adaptive capacity

because it represents the ability of a country to adjust through

substitution of goods, such as alternate protein sources,

production technologies, and accessibility of food resources. A

country with more flexibility in production, trade and

livelihoods is expected to adapt better than countries with

low flexibility.

Flexibility was estimated directly as a country’s trade

balance as reported by the World Bank and OECD national

accounts data, standardized by its average GDP between 1995

and 2005. A greater trade balance indicates greater flexibility to

invest in food imports if domestic production is insufficient or

to adapt to a reduction in exports resulting from a decline in

the production of reef fisheries marketed internationally.

The GINI index, a measure of income inequality, was also

used as an additional, indirect measure of a country’s

flexibility. We assumed that countries with high inequality

exhibit a greater difficulty in shifting resources to support the

more vulnerable populations at the lower end of the income

distribution. Conversely, we assumed that countries with

lower GINI coefficients exhibit a social-political context in

which economic opportunities and burdens are shared more

equally and more people would have access to alternative food

sources in the face of coral reef fisheries decline.

3.3.3. LearningLearning was defined as a country’s organizational and

institutional capacity to access and act on information.

Learning is a component of adaptive capacity because it

allows a country to recognize and respond appropriately to

environmental changes that affect its food systems. A country

that is adapting well would be able to acquire, synthesize, and

incorporate new knowledge into decision making, including

knowledge from resource users. Therefore, two of the most

critical components of learning are the use of science in

fisheries management and the education level of the popula-

tion.

The degree to which countries use science in decision

making for fisheries was derived from the survey dataset

collected from representative national officials by Mora et al.

(2009) between 2007 and 2009. Officials were asked to report on

the education level of fisheries managers, the use of holistic

fisheries models, environmental and biological data, the

frequency of assessments, and the extent to which precau-

tionary measures were used in decision making. A higher

score on this scientific robustness scale indicates a greater

capacity for learning.

We also measured learning as the average adult literacy

rate (15 years and older) in a country between 1995 and 2005 as

reported by the World Bank. Declines in literacy rates may

occur in poor countries when governments reduce spending

on education and health. We assume that countries with

higher adult literacy rates have more resources dedicated to

education; their populations will be better equipped to

respond to changes to food systems, access alternative food

sources, and incorporate new knowledge that will improve

their adaptive capacity.

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8 99

3.3.4. Social organizationSocial organization was defined as the degree to which a

country’s institutional and policy frameworks support or hinder

food security. Social organization is a component of adaptive

capacity because it indicates the ability of a country to

effectively take steps toward change and implement policies

and programs that will lead to successful adaptation. We

estimated social organization using both a national measure of

overall governance and two specific fisheries-related measures.

The Government Effectiveness Index, developed by the

World Bank, was used as a metric for overall governance. It

was derived based on third party surveys of businesses,

households, and NGOs. Using these index values allows us to

gain a relative understanding of government effectiveness in

our target countries (Brewer et al., 2007).

For fisheries-related measures we used the quality of

fisheries management and incorporation of fisheries in plan-

ning documents in a country. The quality of fisheries manage-

ment was estimated using the average scores from the surveys

conducted by Mora et al. (2009) for a country’s policy

transparency, implementation, use of subsidies, and presence

of foreign fishing; the assumption was that countries with

higher quality fisheries management will be more able to adapt

their management systems to change in resource productivity.

Government effectiveness as it pertains to fisheries

management and policy was also examined by assessing

whether governments have been responding to identified

external threats to reefs and fisheries. Specifically, we used a

content analysis method to evaluate the degree to which

fisheries are incorporated into national-level planning pro-

cesses for development and climate change adaptation, based

on a methodology developed by Thorpe et al. (2006) and Ekbom

and Bojo (1997). We began by selecting a national-level

government planning document from each country that

was one of the following in order of preference: a climate-

related adaptation plan, a national development plan, or other

related plan such as a disaster preparedness and response

plan. We then analyzed the documents according to four

diagnostic categories: (1) the mention of fisheries-related

issues, (2) the acknowledgement of the causal linkages

between fisheries-related issues and poverty-related issues,

Table 1 – Summary of adaptive capacity indicators used in th

Category Indicator

Assets Percentage of population with

access to sanitation

Physical

Assets GDP per capita Financia

Assets Reef area per capita Natural

Flexibility Trade balance standardized

by GDP per capita

Ability t

in food

Flexibility GINI index Income

Learning Scientific robustness Use of s

in fisher

Learning Adult literacy rate General

Social Organization Policy transparency, implementation,

use of subsidies, foreign fishing

Overall

manage

Social Organization Government effectiveness index Overall

Social Organization Score indicating mention of fisheries

management in national-level

policy documents

Governm

fisheries

(3) the specification of fisheries-related responses and actions,

and (4) the linkage between the document formulation process

and fisheries related policy and planning processes. For each

diagnostic, a score between 0 and 3 was given: 0 = no mention,

1 = mentioned, but not elaborated upon, 2 = elaborated, 3 = best

practice. Scores for each of the four diagnostic categories were

summed in order to provide a score out of 12 possible points.

Countries with a high value of government effectiveness and a

high score for fisheries policy are expected to have greater social

organization pertaining to fisheries management, to develop

better responses to coral reef fisheries declines, and to have

greater success in implementing measures to prevent and/or

mitigate these declines (Table 1).

3.4. Country selection

Our aim was to generate a list of countries with coral reef

fisheries and available data for each of our vulnerability

indicator components. In our first round of evaluations, we

selected all countries that had coral coverage based on the

UNEP World Conservation Monitoring Centre for a total of 86

countries. This list was reduced to 27 countries based on the

availability of the above adaptive capacity indicator data for

the years between 1995 and 2005. This process allowed us to

evaluate the vulnerability of countries in a theoretically

rigorous way but tended to favor large and organized countries

that collate and submit these data.

3.5. Calculating vulnerability

The index values were standardized based on maximum

values in our dataset and placed on a scale of zero to one using

the following conversion: (X � Xmin)/(Xmax � Xmin). In this way,

the maximum value within each index was set to a relative

value of one. Vulnerability was calculated as (Exposure + -

Sensitivity) � Adaptive Capacity (Fig. 1), so that lower scores

indicate lower levels of vulnerability. The vulnerability scores

were in turn standardized so that zero is the lowest possible

vulnerability score and three is the highest possible

vulnerability score. By standardizing within our analyses,

these calculated values were relative and only meaningful

is study.

Measures Data Source

infrastructure World Bank

l assets IMF World Economic Outlook Database

assets UNEP Coral Reef Atlas

o invest

imports

World Bank and OECD national accounts

inequality World Bank

cientific information

ies policy

Mora et al. (2009)

education level World Bank

quality of fisheries

ment

Mora et al. (2009)

quality of governance World Bank

ent effectiveness in

management

Content analysis of policy documents

(Thorpe et al., 2006; Ekbom and Bojo, 1997)

Fig. 1 – Structure of the fisheries and food security-specific vulnerability index, composite sub-indices, and components

indicators.

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8100

as they relate to this set of countries and data. The mean value

for all countries will be equal to zero and countries with larger

positive values will have the highest relative vulnerability. We

calculated the percent of vulnerability explained by each

variable (exposure, sensitivity, adaptive capacity) using Princi-

pal Component Analysis (PCA) (e.g., Jolliffe, 1986).

We chose to treat each component and the various metrics

used to measure each component equally. We did not use any

weighting system. This choice was deliberate on the part of

the authors as weights reflect a value system specific to a given

policy context. Our goal in this paper is not to guide country

specific polices but rather to demonstrate the creation and

application of a vulnerability index. Policy makers and other

researchers who wish to apply the index of course may wish to

weight the various components or indicators based on the

specific priorities of the decision being made (e.g., McClanahan

et al., 2008). Those who do decide to employ weights should

consult the Multi-Objective Decision Analysis literature for

guidance on soliciting and incorporating weighted criteria

into the decision process (Hammond et al., 1999; Saaty, 1986).6

6 We thank an anonymous reviewer for highlighting this for us.

3.6. Comparing approaches

To assess the relative gains of applying a food security

perspective to national-scale vulnerability assessments, our

results were compared to those from the World Resource

Institute’s most recent Reefs at Risk report, which assessed the

general vulnerability of countries with coral reefs to the threat

of reef loss due to coastal development, watershed-based

pollution, marine-based pollution, overfishing, thermal stress

and ocean acidification using a wide range of data sources and

methods (Burke et al., 2011). The resulting threat levels for the

reefs in each country were ultimately categorized as low,

medium, or high, with those reefs classified as medium or high

considered ‘‘threatened.’’ Measures of adaptive capacity and

dependence on reefs are then included in a vulnerability

assessment of the social and economic implications of reef loss.

Reef dependence is measured as the proportion of the

population that lives near reefs or fish reefs, the country’s

reef-associated exports, nutritional dependence on fish, reef-

associated tourism, and shoreline protection. Adaptive capacity

is measured as economic resources, education, health, gover-

nance (including fisheries subsidies), access to markets, and

agricultural resources.

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8 101

4. Results

Country-specific vulnerability scores ranged from 0 to 2.33

(Table 2). Single variable regression analyses indicated that

adaptive capacity had a greater influence on vulnerability

compared to exposure and sensitivity. PCA revealed that

exposure, sensitivity, and adaptive capacity explained 39%,

34% and 27% of the variation in vulnerability scores,

respectively. The exposure values were driven largely by

Indonesia and the Philippines, which had very high exposure

scores (1.00 and 0.88, respectively) compared to all the other

countries, which had exposure scores below 0.20.

The component scores for each country indicated that

there was no single driver of vulnerability or a single

underlying mechanism that makes a country particularly

vulnerable to declines in coral reef fisheries; rather, countries

experienced vulnerability as the result of a unique combina-

tion of adaptive capacity, sensitivity and exposure (Fig. 2). In

this way we identified countries whose particular combination

of factors warranted greater attention or provided an

interesting combination of factors for further evaluation.

For example, Indonesia and Liberia were the most vulnerable

countries overall, but the drivers of their high levels of

vulnerability were different: Indonesia had the highest

exposure levels while Liberia’s vulnerability resulted from

very high sensitivity to coral reef fisheries decline and the

lowest level of adaptive capacity. Liberia’s sensitivity score

Table 2 – Standardized exposure, sensitivity, adaptive capacitsure + sensitivity) S adaptive capacity.a

Country Exposure Sensitivi

Indonesia 1.00 0.98

Liberia 0.0050 0.92

Ivory Coast 0.00025 0.91

Kenya 0.016 0.81

Philippines 0.88 0.40

Honduras 0.019 0.90

Cameroon 0.00044 0.91

Egypt 0.10 0.87

Cambodia 0.00051 0.90

Tanzania 0.076 0.94

Bangladesh 0.00 0.88

Comoros 0.0067 0.74

Nicaragua 0.012 0.89

Cape Verde 0.0065 0.86

India 0.084 0.44

Senegal 0.0015 0.66

Madagascar 0.037 0.21

China 0.025 0.78

Brazil 0.019 0.91

Costa Rica 0.014 1.00

Panama 0.025 0.78

Mexico 0.038 0.88

Trinidad and Tobago 0.0015 0.93

Thailand 0.053 0.85

Dominican Republic 0.016 0.39

Sri Lanka 0.019 0.12

Malaysia 0.083 0.00

Mean 0.094 0.74

a Values for each variable have been standardized individually.

was driven by its close proximity to the 30 g/capita/day

threshold recommended by the FAO and WHO; Liberia’s

average daily protein intake was 34.5 g, the lowest of any of our

studied countries. Unlike Liberia, other countries with the

lowest vulnerability scores, such as Sri Lanka, Malaysia, and

the Dominican Republic, had high adaptive capacity scores

relative to their levels of exposure and sensitivity.

A number of African countries, including Kenya, Camer-

oon, Ivory Coast, and Comoros, had higher than average

vulnerability scores despite relatively small coral reef areas.

These countries had very low adaptive capacity scores and

higher than average levels of sensitivity to declines in coral

reef fisheries production. Another category of vulnerability

was found in Latin American countries such as Brazil, Costa

Rica, and Mexico, which had high adaptive capacity but very

high sensitivity, which resulted in slightly below-average

vulnerability scores.

As described above, adaptive capacity played the largest

role in determining vulnerability scores. Further analysis of

the indicators used in the adaptive capacity scores reveals

patterns among the countries (Table 3). First, there is not much

variation in the social organization scores, which ranged from

0.75 to 1.99 out of a total possible score of 3. Flexibility scores

were consistently low for the 27 countries (with possible

scores ranging from �1 to 1), and often less than zero due to

the consistently poor GINI scores. The differences in adaptive

capacity, therefore, are driven largely by differences in assets

and learning. For example, the country with the lowest

y, and vulnerability scores where vulnerability = (expo-

ty Adaptive capacity Vulnerability

0.37 2.33

0.00 1.65

0.15 1.48

0.10 1.45

0.57 1.43

0.23 1.41

0.25 1.37

0.45 1.24

0.43 1.19

0.55 1.19

0.46 1.14

0.34 1.12

0.51 1.12

0.50 1.08

0.31 0.94

0.47 0.92

0.063 0.91

0.65 0.87

0.78 0.86

0.89 0.85

0.71 0.82

0.88 0.75

0.91 0.74

1.00 0.62

0.68 0.45

0.84 0.023

0.80 0.00

0.51 1.04

Fig. 2 – Ternary plot showing the unique combination of factors underlying individual countries’ vulnerability scores.

Table 3 – Scores for the components of adaptive capacity: assets, flexibility, learning and social organization.

Country Assets Flexibility Learning Social organization Adaptive capacity total

Liberia 0.058 �0.36 0.44 1.20 1.34

Madagascar 0.42 0.11 0.28 0.75 1.57

Kenya 0.31 0.041 0.50 0.84 1.70

Ivory Coast 0.23 0.54 0.22 0.89 1.89

Honduras 1.01 �0.25 0.50 0.93 2.19

Cameroon 0.52 0.28 0.31 1.15 2.27

India 0.23 0.47 0.16 1.60 2.46

Comoros 1.26 �0.70 0.70 1.34 2.60

Indonesia 0.78 0.55 0.19 1.16 2.68

Cambodia 0.11 0.13 1.24 1.42 2.90

Egypt 1.19 0.54 0.16 1.08 2.97

Bangladesh 0.45 0.59 0.00 1.99 3.03

Senegal 0.47 0.11 1.41 1.07 3.06

Cape Verde 0.68 �0.60 1.40 1.71 3.18

Nicaragua 0.81 �0.48 1.83 1.05 3.20

Tanzania 0.34 0.38 1.24 1.39 3.34

Philippines 1.26 0.18 0.31 1.66 3.41

China 0.59 0.42 1.25 1.48 3.74

Dom. Rep. 1.29 0.0053 1.00 1.53 3.83

Panama 1.79 0.028 0.57 1.54 3.93

Brazil 1.33 �0.12 1.89 1.09 4.20

Malaysia 1.78 0.53 0.34 1.63 4.28

Sri Lanka 0.97 0.35 1.14 1.94 4.40

Mexico 1.70 0.11 1.60 1.17 4.58

Costa Rica 1.59 0.15 1.58 1.27 4.59

Trin. & Tob. 2.02 0.60 0.25 1.80 4.67

Thailand 1.29 0.40 1.34 1.97 5.00

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8102

adaptive capacity score, Liberia, has a score of 0.058 (out of 3)

for assets and 0.44 (out of 2) for learning while the country with

the highest adaptive capacity score, Thailand, has scores of

1.29 and 1.34, respectively.

4.1. Comparing approaches

The food security-oriented and the more general Reefs at Risk

vulnerability assessments produced some similarities and

Table 4 – Comparison of two vulnerability measures: afood security approach and a more general vulnerabilityassessment (Burke et al., 2011). Countries scaled frommost to least vulnerable by the food-security approach.

Country Vulnerability:food security

approach

Vulnerability:general approach(Burke et al., 2011)

Indonesia 2.33 Very High

Liberia 1.65 –

Ivory Coast 1.48 –

Kenya 1.45 High

Philippines 1.43 Very High

Honduras 1.41 Medium

Cameroon 1.37 –

Egypt 1.24 Medium

Cambodia 1.19 Medium

Tanzania 1.19 Very High

Bangladesh 1.14 Low

Comoros 1.12 Very High

Nicaragua 1.12 Low

Cape Verde 1.08 –

India 0.94 Medium

Senegal 0.92 –

Madagascar 0.91 Very High

China 0.87 Medium

Brazil 0.86 Low

Costa Rica 0.85 Medium

Panama 0.82 High

Mexico 0.75 Low

Trinidad and Tobago 0.74 Medium

Thailand 0.62 High

Dominican Republic 0.45 Very High

Sri Lanka 0.023 High

Malaysia 0.00 Medium

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8 103

differences in the scaling of national vulnerability (Table 4).

There was agreement between the two methodologies that

Indonesia, Kenya and the Philippines were very vulnerable

countries. However, this was where the agreement ended.

Those countries that were the least vulnerable from a food

security perspective (i.e., Sri Lanka, Dominican Republic, and

Thailand) were considered to be highly vulnerable using the

more general assessment. In addition, countries that had near

average vulnerability scores using the food security specific

metrics had general vulnerability scores that ranged from low

to very high (i.e., Nicaragua and Madagascar). In summary, the

most vulnerable countries from a food security perspective are

also highly vulnerable using a more general vulnerability

assessment, but the two methodologies identify different sets

of countries as having low and medium range levels of

vulnerability.

5. Discussion

The aim of our study was to develop a sector scale-specific,

policy relevant vulnerability assessment framework focused

on food insecurity implications of disturbances to coral reef

ecosystems and their associated fisheries. We drew on a range

of data sources and theoretical underpinnings to create a

vulnerability index based on adaptive capacity, exposure, and

sensitivity. We found a considerable range of adaptive

capacity, exposure, and sensitivity among the countries

studied and, therefore, differences in the overall food security

vulnerability to a decline in reef fisheries. From a food security

perspective, and of the countries we examined, the most

vulnerable countries to coral reef fisheries decline are

Indonesia and Liberia while the least vulnerable countries

are Sri Lanka, Malaysia, and the Dominican Republic. Each

country in our analysis has a unique suite of factors

underlying its vulnerability scores, particularly those coun-

tries at the higher end of the vulnerability spectrum.

Adaptive capacity had the greatest effect on vulnerability

scores, explaining 39% of their variation. This suggests there is

considerable opportunity to influence vulnerability through

interventions that build adaptive capacity. Flexibility mecha-

nisms are lacking in the majority of countries due to high

levels of income inequality and exposure of national econo-

mies to trade deficits. These aspects of tropical country

economies may create challenges in their ability to adapt to

the expected increases of climate change impacts and areas of

policy needing immediate examination and possible reforms.

Further, boosting the country’s critical assets – financial,

physical, and natural – and the learning capabilities of the

population and government is expected to greatly assist the

adaptation potential of countries with low adaptive capacity.

In many African countries, such as Kenya, Cameroon, Ivory

Coast, and Comoros, low adaptive capacity explains higher

than average vulnerability. In these countries, policy and

development efforts to promote food security may benefit

most from a primary focus on developing assets and flexibility.

Actions may include increasing the amount of a country’s

reefs that is included in marine protected areas, developing

other fisheries restrictions to help rebuild fish stocks,

increasing alternate protein sources through agriculture and

aquaculture, reducing income inequality through pro-poor

and fair labor policies and practices, and boosting fair trade

and green markets.

Some countries, such as Liberia and Honduras, were

classified as vulnerable due to the combination of high levels

of food security sensitivity to coral reef fisheries decline and

low levels of adaptive capacity. This combination of scores

indicates that these countries are likely to be highly sensitive

and will struggle to adapt if environmental conditions were to

worsen – for example if fish stocks were to decline or ocean

acidification to increase – the country’s people would be most

sensitive and would find it very challenging to adapt. For

example, while Liberia has a relatively small reef area, its low

levels of protein intake (as reported by the FAO) contribute to

its people’s vulnerability from a food security perspective. In

such cases, policy interventions should focus first on reducing

sensitivity and second on increasing adaptive capacity. This

specifically means finding ways to increase sources of protein

and nutrient-dense foods in these countries and the social

mechanisms to make this sustainable.

A third combination of features with policy implications is

found in middle income countries such as Thailand, Costa

Rica, and Mexico that have high adaptive capacity scores – and

thus relatively low vulnerability scores – but high levels of

sensitivity. Based on present conditions, their high levels of

adaptive capacity are compensating for the sensitivity of their

populations to changes in coral reef fisheries productivity.

7 A full lists of Small Island Developing states used in thisanalysis is available from the United Nations at: http://www.un.org/special-rep/ohrlls/sid/list.htm.

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8104

However, if conditions in these countries were to deteriorate

in any way – if exposure increased due to fishing pressures

that exceeds maximum yields or adaptive capacity decreased

due to political conflict or natural disasters – their high

sensitivity indicates that these changes would have signifi-

cant consequences; policies to reduce this sensitivity could

then be promoted. This would mean leveraging their

relatively high levels of adaptive capacity to quickly increase

the availability and accessibility of alternative protein

sources (McClanahan et al., 2009). If fisheries or ocean

conditions were to change slightly, or if political institutions

were to erode, the high sensitivity of these countries means

their vulnerability could increase greatly. Allison (2011)

argues these medium-level developed countries may be

among the best locations to develop small to medium size

enterprise aquaculture.

In comparing our findings to Burke’s et al. (2011) more

general approach to evaluating vulnerability to coral reef

decline, we find that our food security-specific metrics identify

somewhat different priorities for intervention, particularly

those for low to medium range generic vulnerability nations.

The differences are due to the fisheries-specific indicators for

sensitivity and adaptive capacity used in our study that provide

information about the implications of coral reef fisheries

decline specifically for the country’s food system. The general

approach would give priority to countries that, from a food

security perspective, would not be prioritized highly, and

include Thailand, the Dominican Republic, and Sri Lanka.

Other countries, such as Honduras and Bangladesh, would be

prioritized highly by a food security but not a general approach.

In either case, policies and resources may be inefficiently

allocated and opportunities to increase food security lost when

the evaluations and actions are not sufficiently specific. This

comparison found differences between the two approaches

that suggest a need for sector and policy-specific diagnostics

when developing vulnerability metrics. The major difference

between the Reefs at Risk approach and this food security

approach is our use of a more specific food security-oriented

metric for sensitivity and a more targeted assessment of

adaptive capacity as it relates specifically to governing fisheries.

The major limitation of this study – and likely of other

efforts to produce national scale, policy-specific vulnerability

metrics – is the difficulty in finding the specific data or

measuring the key features of vulnerability at the national

scale. This is a common problem to overcome when using

indicators (Langbein and Knack, 2010; Birkmann, 2007) but the

data limitations in this case were particularly severe. Of the 86

countries reporting reef fisheries catch, we were only able to

gather sufficient data for 27 even after an extensive search for

suitable indicators. The data deficiency was particularly

prominent for Small Island Developing States (SIDS), countries

that are particularly dependent on coral reefs for food (Thorpe

et al., 2005; Bell et al., 2009) and for whom this type of analysis

may be particularly valuable. We emphasize that our

vulnerability scores are relative and unique to our country

list. Some metrics are more readily available, but slightly less

theoretically relevant and robust, but could be substituted into

our framework (see Appendix A). A challenge to data

availability arises from the definition of a nation: islands that

are often most dependent on coral reef fisheries are commonly

territories of other nations, and data unique to such islands

are not available in large databases, such as those maintained

by the FAO. However, data from island territories may be

available from local government agencies or non-governmen-

tal organizations and we encourage investigators and policy

makers to find and use metrics that are available for their

countries. Improving national data sets on infrastructure,

governance, food systems, and fisheries is imperative to

further efforts to understand national scale vulnerability.

6. Conclusions

The approach to policy-specific, national-scale vulnerability

assessment developed in this paper is valuable for conceptu-

alizing key factors influencing national vulnerability,

expanding existing frameworks and tools, and prioritizing

policy needs and actions associated with food security

problems. Future investigations in poor and particularly

African countries should examine the factors that best

promote adaptive capacity to manage coral reef fisheries

and prevent declines in their food production. In middle-

income countries the focus should be on factors that best

reduce their sensitivities; put simply, such countries have

alternatives to eating their coral reef fish populations and

efforts to develop them should be a priority. Better data are

needed if these types of assessments are to expand

geographically, particularly to include vulnerable SIDS

countries, and to other factors beyond the fisheries sector.

The food security of many countries will be undermined by

declining coral reef fisheries resources and using a scale and

sector-specific vulnerability should help identify the key

constraints and the most useful actions for reducing them.

Acknowledgements

The authors thank Alice Alldredge, Ben Halpern, Chris Funk,

Stuart Sweeney, and James Watson for their input and

assistance. The Luce and the John D. and Catherine T.

McArthur Foundations provided financial support and the

National Center for Ecological Applications and Synthesis

provided computing and additional resources.

Appendix A. Revising the framework for SIDS

Small Island Developing States (SIDS) are of special concern

for coral reef fishery management and food security (Thorpe

et al., 2005; Ghina, 2003; Bell et al., 2009). Many SIDS have an

exceptionally high dependence on the fish and shellfish

associated with coral reefs for food. However, due to the lack

of available and consistent country-level data, many SIDS

were not included in our assessment of food security

vulnerability to coral reef fisheries decline. There are 52

countries classified by the United Nations as SIDS7 but only 4

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8 105

of the 27 countries analyzed in this paper are SIDS due to data

constraints. This indicates that better data are needed for

these countries and, absent such data improvements, an

alternative vulnerability index is required to assess the food

security vulnerability of the vast majority of SIDS. This

appendix is included to determine the components of our

vulnerability index that were most data-deficient for SIDS and

to propose a scaled-down version that replaces some of the

metrics with more readily available but less theoretically

robust data alternatives.

Exposure and sensitivity data were available for 35 SIDS. A

major impediment to exposure and sensitivity data availabili-

ty for these countries is the fact that major data collection

centers such as the FAO and UNEP do not collect and store data

for these countries individually and instead use broader

groupings of countries. For example, the FAO does not collect

country-level fisheries and nutrition data for Tuvalu, Nauru, or

Palau. Exposure data are often collected and analyzed at

spatial resolutions that are incompatible with the size of the

smallest SIDS. In the case of exposure and sensitivity data,

therefore, values could be interpolated from nearby areas or

produced through local ground trothing and data collection

efforts.

In our analysis, the majority of SIDS were excluded due to a

lack of adaptive capacity indicators, in particular the GINI

index (available for 19 SIDS) and the adult literacy rate

(available for 21 SIDS). Two other adaptive capacity indicators,

trade balance standardized by GDP (available for 37 SIDS) and

the percentage of the population with access to sanitation

(available for 36 SIDS) were somewhat limiting for SIDS, but

less problematic. The fisheries-specific measures from Mora

et al. (2009) (policy transparency, implementation, use of

subsidies, foreign fishing under the social organization

category, and scientific robustness under the learning catego-

ry) tended to have good coverage for SIDS, likely because this

was a fisheries specific study. The government effectiveness

index component of social organization, developed by the

World Bank, was available for 48 SIDS. The coral reef area was

obtained from the same source used by Halpern et al. (2008)

and is therefore available for all countries that had exposure

data. Finally, policy documents used in the analysis following

methods of Thorpe et al. (2006) were found for all of the SIDS

Table A1 – A revised vulnerability framework for food security aavailable but less theoretically robust indicators that would im

Category Indicator

Assets Percentage of population with access to sanitation

Assets GDP per capita

Assets Coral reef area normalized by population

Flexibility Trade balance standardized

by GDP per capita

Flexibility Life expectancy at birth

Learning Scientific robustness

Learning Primary education duration

Social Organization Policy transparency, implementation,

use of subsidies, foreign fishing

Social Organization Government effectiveness index

that had data for the other adaptive capacity indicators and is

not likely to be a limiting factor.

As detailed above, the measures that had the worst

coverage for SIDS nations were national level measures

available from the World Bank (the GINI index and adult

literacy). Replacing these two limiting adaptive capacity

indicators with cruder proxies would allow for greater SIDS

coverage.

To construct a more data rich version of the index, we first

replaced the GINI index with life expectancy at birth, which

has been argued to be strongly correlated with social

inequality as well as total measures of wealth (Riley, 2005).

Additionally, this measure is a component of the Human

Development Index and was specifically selected to capture

notions of human development not captured by GDP, which

include inequality (http://hdr.undp.org/en/statistics/hdi/) and

was utilized by the Reefs at Risk Revisited report as one of 6

indicators combined to represent adaptive capacity (Burke

et al., 2011), and by Bell et al. (2011) as part of their health

component of adaptive capacity. Life expectancy at birth is

available from the World Bank for 42 SIDS countries, and while

it is an indirect measure of inequality may serve as a useful

proxy for SIDS.

Secondly, we replaced adult literacy rate with primary

education duration in years, which is also available from the

World Bank for 44 SIDS countries. This metric was used by Bell

et al. (2001) along with literacy to assess their education

component of adaptive capacity. Primary education duration

is again similar to the measure used by the HDI, which uses a

(slightly different) measure of years of schooling as a proxy for

education levels. Again, while primary education duration is

an indirect measure of literacy, a critical component for

learning and flexibility, it may serve as a useful proxy for SIDS.

Table A1 presents a set of variables that includes these

changes, and we believe that this new index would provide

similar intellectual framework while encompassing signifi-

cantly more SIDS nations. Given that the final version includes

two measures of Learning, Flexibility, and Social Organization,

and three measures of Assets, removing percent of population

with access to sanitation (the next limiting factor), may

provide coverage for a more countries while still maintaining a

balanced framework (it has been italicized to indicate this).

nd coral reef fisheries decline that exchanges more readilyprove coverage of the framework for SIDS.

Measures Data Source

Physical infrastructure World Bank

Financial assets World Bank

Natural assets UNEP

Ability to invest

in food imports

World Bank and

OECD national accounts

Income inequality World Bank

Use of scientific information

in fisheries policy

Mora et al. (2009)

General education level World Bank

Overall quality of fisheries

management

Mora et al. (2009)

Overall quality of governance World Bank

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8106

r e f e r e n c e s

Adger, W.N., 2006. Vulnerability. Global Environmental Change16 (3), 268–281, http://dx.doi.org/10.1016/j.gloenvcha.2006.02.006.

Adger, W.N., Kelly, P.M., 1999. Social vulnerability to climatechange and the architecture of entitlements. Mitigation andAdaptation Strategies for Global Change 4, 253–266.

Allison, E.H. 2011. Aquaculture, Fisheries, Poverty and FoodSecurity. Working Paper. Penang, Malaysia: World FishCenter.

Allison, E.H., Perry, A.L., Badjeck, M.-C., Neil Adger, W., KatrinaBrown, Declan Conway, Halls, A.S., et al., 2009. Vulnerabilityof national economies to the impacts of climate change onfisheries. Fish and Fisheries 10, 173–196, http://dx.doi.org/10.1111/j.1467-2979.2008.00310.x.

Atta-Mills, J., Alder, J., Sumalia, R.U., 2004. The decline of aregional fishing nation: the case of Ghana and West Africa.Natural Resources Forum 28 (1), 13–21, http://dx.doi.org/10.1111/j.0165-0203.2004.00068.x.

Bell, J.D., Johnson, J.E., Hobday, A.J., 2011. Vulnerability ofTropical Pacific Fisheries and Aquaculture to ClimateChange. Secretariat of the Pacific Community, NewCaledonia.

Bell, J.D., Kronen, M., Vunisea, A., Nash, W.J., Keeble, G.,Demmke, A., Pontifex, S., Andrefouet, S., 2009. Planning theuse of fish for food security in the Pacific. Marine Policy 33(January (1)), 64–76, http://dx.doi.org/10.1016/j.marpol.2008.04.002.

Birkmann, J., 2007. Risk and vulnerability indicators at differentscales: applicability, usefulness and policy implications.Environmental Hazards 7 (1), 20–31, http://dx.doi.org/10.1016/j.envhaz.2007.04.002.

Brewer, G.A., Yujin Choi, Walker, R.M., 2007. Accountability,corruption and government effectiveness in Asia: anexploration of World Bank governance indicators.International Public Management Review 8 (2), 200–219.

Brooks, N.W., Neil Adger, P., Mick Kelly, 2005. The determinantsof vulnerability and adaptive capacity at the national leveland the implications for adaptation. Global EnvironmentalChange Part A 15 (2), 151–163, http://dx.doi.org/10.1016/j.gloenvcha.2004.12.006.

Brown, M.E., Funk, C.C., 2008. Food security under climatechange. Science 319 (February (5863)), 580–581, http://dx.doi.org/10.1126/science.1154102.

Burke, L., Maidens, J., 2004. Reefs at Risk in the Caribbean. WorldResources Institute, Washington, DC.

Burke, L., Katie Reytar, Mark Spalding, Perry, A.L., 2011. Reefs atRisk Revisited. World Resources Institute, Washington, DC.

Burke, L., Selig, E., Spalding, M., 2002. Reefs at Risk in SoutheastAsia. World Resources Institute, Washington, DC.

Cinner, J.E., Fuentes, M.M.P.B., Randriamahazo, H., 2009.Exploring social resilience in madagascar’s marine protectedareas. Ecology and Society 14 (1), 41.

Cinner, J.E., McClanahan, T.R., 2006. Socioeconomic factors thatlead to overfishing in small-scale reef fisheries in Papua NewGuinea. Environmental Conservation 33, 73–80.

Cinner, J.E., McClanahan, T.R., Graham, N.A.J., Daw, T.M.,Maina, J., Stead, S.M., Wamukota, A. Brown, K, Bodin, O.Vulnerability of coastal communities to key impacts ofclimate change on coral reef fisheries. Global EnvironmentalChange (0). doi:10.1016/j.gloenvcha.2011.09.018. http://www.sciencedirect.com/science/article/pii/S0959378011001579.

Dey, M.M., Rab, M.A., Paraguas, F.J., Somying Piumsombun,Ramachandra Bhatta, Md Ferdous Alam, MahfuzuddinAhmed, 2005. Fish consumption and food security: a

disaggregated analysis by types of fish and classes ofconsumers in selected Asian countries. AquacultureEconomics & Management 9 (1–2), 89–111, http://dx.doi.org/10.1080/13657300590961537.

Dilley, M., Boudreau, T.E., 2001. Coming to terms withvulnerability: a critique of the food security definition. FoodPolicy 26 (June (3)), 229–247, http://dx.doi.org/10.1016/S0306-9192(00)00046-4.

Dulvy, N., Allison, E., 2009. ‘‘A Place at the Table?’’ (0906) (June):68-70. doi:10.1038/climate.2009.52.

Ekbom, A., Bojo, J., 1997. Mainstreaming Environment inCountry Assistance Strategies. Discussion Paper. WorldBank, Washington, D.C.

FAO, 2011. The State of Food Insecurity in the World 2010.United Nations Food and Agriculture Organization, Rome.

Ghina, F., 2003. Sustainable development in small islanddeveloping states. Environment, Development andSustainability 5 (1–2), 139–165.

Gregory, P.J., Ingram, J.S.I., Brklacich, M., 2005. Climate changeand food security. Philosophical Transactions of the RoyalSociety B: Biological Sciences 360, 2139–2148.

Halpern, B.S., Walbridge, S., Selkoe, K.A., Kappel, C.V., FiorenzaMicheli, Caterina D’Agrosa, Bruno, J.F., et al., 2008. A globalmap of human impact on marine ecosystems. Science 319(February (5865)), 948–952, http://dx.doi.org/10.1126/science.1149345.

Hammond, J.S., Ralph Keeney, Howard Raiffa, 1999. SmartChoices: A Practical Guide to Making Better Decisions.Harvard University Press, Cambridge, MA.

Handmer, J.W., Dovers, S., Downing, T.E., 1999. Societalvulnerability to climate change and variability. Mitigationand Adaptation Strategies for Global Change 4, 267–281.

Hicks, C.C., McClanahan, T.R., Cinner, J.E., Hills, J.M., 2009.Trade-offs in values assigned to ecological goods andservices associated with different coral reef managementstrategies. Ecology and Society 14 (1), 10.

Ionescu, C., Klein, R.J.T., Jochen Hinkel, Kavi Kumar, K.S., RupertKlein, 2009. Towards a formal framework of vulnerability toclimate change. Environmental Modeling and Assessment 14(1), 1–16.

Jolliffe, I.T., 1986. Principal Component Analysis. Springer-Verlag, New York.

Kawarazuka, N., Bene, C., 2010. Linking small-scale fisheriesand aquaculture to household nutritional security: anoverview. Food Security 2 (4), 343–357.

Kawarazuka, N., Bene, C., 2011. The potential role of small fishspecies in improving micronutrient deficiencies indeveloping countries: building the evidence. Public HealthNutrition 14, 1927–1938.

Kent, G., 1997. Fisheries, food security, and the poor. Food Policy22 (October (5)), 393–404, http://dx.doi.org/10.1016/S0306-9192(97)00030-4.

Langbein, L., Knack, S., 2010. The worldwide governanceindicators: six, one, or none? Journal of DevelopmentStudies 46 (2), 350–370.

Leurs, A.L., 2005. The surface of vulnerability: an analyticalframework for examining environmental change. GlobalEnvironmental Change 15, 214–223.

Luers, A.L., Lobell, D.B., Sklar, L.S., Addams, C.L., Matson, P.A.,2003. A method for quantifying vulnerability, applied to theagricultural system of the Yaqui Valley, Mexico. GlobalEnvironmental Change 13 (December (4)), 255–267, http://dx.doi.org/10.1016/S0959-3780(03)00054-2.

McClanahan, T.R., 2002. The near future of coral reefs.Environmental Conservation 29, 460–483.

McClanahan, T.R., Cinner, J.E., Graham, N.A.J., Daw, T.M.,Maina, J., Stead, S.M., Wamukota, A., Katrina Brown, Venus,V., Polunin, N.V.C., 2009. Identifying reefs of hope andhopeful actions: contextualizing environmental, ecological,

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8 107

and social parameters to respond effectively to climatechange. Conservation Biology 23 (3), 662–671.

McClanahan, T.R., Cinner, J.E., Maina, J., Graham, N.A.J., Daw,T.M., Stead, S.M., Wamukota, A., et al., 2008. Conservationaction in a changing climate. Conservation Letters 1 (2),53–59, http://dx.doi.org/10.1111/j.1755-263X.2008.00008_1.x.

Mora, C., Myers, R.A., Coll, M., Libralato, S., Pitcher, T.J., 2009.Management effectiveness of the World’s marine fisheries.PLoS Biology 7 (6), e1000131.

Newton, K., Cote, I.M., Pilling, G.M., Simon Jennings, Dulvy, N.K.,2007. Current and future sustainability of island coral reeffisheries. Current Biology 17 (7), 655–658.

Pandolfi, J.M., Connolly, S.R., Marshall, D.J., Cohen, A.L., 2011.Protecting coral reef futures under global warming andocean acidification. Science 333, 418–422.

Pauly, D., Watson, R., Alder, J., 2005. Global Trends in WorldFisheries: Impacts on Marine Ecosystems and Food Security.Philosophical Transactions of the Royal Society B: BiologicalSciences 360 (January (1453)), 5–12, http://dx.doi.org/10.1098/rstb.2004.1574.

Pelling, M., Uitto, J.I., 2001. Small island developing states:natural disaster vulnerability and global change.Environmental Hazards 3, 49–62.

Riley, J.C., 2005. Poverty and Life Expectancy. CambridgeUniversity Press, Cambridge, UK.

Saaty, Thomas L., 1986. Decision Making for Leaders: TheAnalytic Hierarchy Process for Decisions in a ComplexWorld. RWS Publications, Pittsburgh, PA.

Schmidhuber, J., Tubiello, F.N., 2007. Global food security underclimate change. Proceedings of the National Academy ofSciences 104 (December (50)), 19703–19708, http://dx.doi.org/10.1073/pnas.0701976104.

Smit, B., Wandel, J., 2006. Adaptation, Adaptive Capacity andVulnerability. Global Environmental Change 16 (August (3)),282–292, http://dx.doi.org/10.1016/j.gloenvcha.2006.03.008.

Smith, M.D., Roheim, C.A., Crowder, L.B., Halpern, B.S.,Turnipseed, M., Anderson, J.L., Asche, F., et al., 2010.Sustainability and global seafood. Science 327 (February(5967)), 784–786, http://dx.doi.org/10.1126/science.1185345.

Smith, S.V., Buddemeier, R.W., 1992. Global change and coralreef ecosystems. Annual Review of Ecology and Systematics23 (1), 89–118.

Thorpe, A., Chris Reid, Raymon Van Anrooy, Cecile Brugere,2005. Integrating Fisheries into the National DevelopmentPlans of Small Island Developing States (SIDS): ten years onfrom Barbados. Natural Resources Forum 29 (1), 51–69, http://dx.doi.org/10.1111/j.1477-8947.2005.00112.x.

Thorpe, A., Chris Reid, Raymon van Anrooy, Cecile Brugere,Denis Becker, 2006. Asian development and povertyreduction strategies: integrating fisheries into thedevelopment discourse. Food Policy 31 (5), 385–400, http://dx.doi.org/10.1016/j.foodpol.2005.09.007.

Turner, B.L., Kasperson, R.E., Matson, P.A., McCarthy, J.J., Corell,R.W., Lindsey Christensen, Noelle Eckley, et al., 2003. Aframework for vulnerability analysis in sustainabilityscience. Proceedings of the National Academy of Sciences100 (14), 8074–8079.

Whittingham, E., Campbell, J., Townsley, P., 2003. Poverty andReefs. DFID-IMM-IOC/UNESCO.

Wilkinson, C.R., Buddemeier, R.W., 1994. Global climate changeand coral reefs: implications for people and reefs. Report ofthe UNEP-IOC-ASPEI-IUCN Global Task Team on theImplications of Climate Change on Coral Reefs. Gland,Switzerland: IUCN.

Sara Hughes is a postdoctoral fellow in the Research ApplicationsLaboratory of the National Center for Atmospheric Research. Herresearch focuses on the role of social and political institutions in

environmental policy and climate change adaptation. Sara re-ceived her PhD in 2011 in Environmental Science and Managementfrom the University of California, Santa Barbara. Her previouswork has examined the politics and institutions engaged in urbanwater management reform and currently Sara’s research is aimedat better understanding the drivers of urban climate change plan-ning and its implications for equity and development in the U.S.,Asia, and Latin America.

Annie Yau earned a Ph.D. in Environmental Science and Man-agement from the University of California, Santa Barbara in 2011.She developed methods to model and manage populations atsmall spatial scales under uncertainty in the amount of self-recruitment, using the example of a giant clam fishery in FrenchPolynesia. She currently advises on National Ocean Policy issuesfor the National Oceanic and Atmospheric Administration. Pre-viously, she evaluated the sustainability of several fisheries forthe Seafood Watch Program of the Monterey Bay Aquarium.She has also published on the photophysiology of symbioticmarine algae and best education practices for environmentaleducation.

Lisa Max is a marine ecology PhD student in the Ecology, Evolutionand Marine Biology Department at the University of California,Santa Barbara, and holds a Master’s degree in EnvironmentalManagement from Yale University, School of Forestry and Envi-ronmental Studies. Lisa’s dissertation focuses on food web dy-namics in coral reef and kelp forest ecosystems.

Nada Petrovic is a Postdoctoral Fellow at the Center for Researchon Environmental Decisions at Columbia University. She is broad-ly interested in how perceptions of environmental issues influ-ence decisions on an individual and community level. Inparticular, she focuses on the interpretation and use of scientificinformation in the context of climate change and natural disas-ters. Her background is in physical science, and in her doctoralwork she studied optimal decision-making for wildfire responseusing a numerical modeling approach.

Frank Davenport is a PhD student in Geography at UCSB. Hisresearch focuses on food security and spatial econometrics. Frankis interested in how trade liberalization impacts spatial pricebehavior among agricultural commodity markets. One of his dis-sertation papers analyzes how the response of Mexican maizeprices to local and global influences varies over space and timeduring different stages of the North American Free Trade Agree-ment (NAFTA). Frank is also working with colleagues at the UCSBClimate Hazards Group to examine maize prices, climate trends,and child malnutrition in Kenya.

Michael Marshall is broadly interested in how coupled land sur-face-atmospheric processes impact agrarian society. He receivedhis Ph.D. in Geography from UC Santa Barbara in 2010. His disser-tation, titled Modeling Evapotranspiration in sub-Saharan Africa: ATool for Food Security Analysis, synthesized remote sensing and landsurface reanalysis to estimate evapotranspiration. He was recent-ly awarded a Mendenhall Research Fellowship through the U.S.Geological Survey to combine ground, hyper- spatial and spectralremote sensing, and ancillary spatial data to estimate and evalu-ate crop water productivity.

Timothy R. McClanahan is a coral reef ecosystem ecologist withresearch interests spanning the fields of marine protected areas,food webs, nutrients, fisheries, climate change, resilience, and thelinkages between coral reef ecosystems and the humans whodepend on them. He has spent most of his professional life livingand working in Kenya, and for the last 20 years has worked as a

e n v i r o n m e n t a l s c i e n c e & p o l i c y 2 3 ( 2 0 1 2 ) 9 5 – 1 0 8108

Senior Conservation Zoologist for the Wildlife Conservation Soci-ety, based in Mombasa, on the east African coast. His work hasfocused on providing solutions to human-coral reef fisheries in-teraction in poor developing countries.

Dr. Edward H. Allison has over 20 years’ experience in the field offisheries management and development in sub-Saharan Africa,Asia, Oceania, Latin America and the UK, as researcher or tech-nical and policy advisor for various international organizations.He currently holds a part-time faculty position at the Universityof East Anglia, U.K, where his research focuses on the contribu-tion of fisheries and aquaculture to food and nutrition security,and coastal and riparian people’s vulnerability and adaptation toclimate change. In 2013 he will take up a professorship in the

School of Marine and Environmental Affairs, University ofWashington.

Dr. Joshua E. Cinner’s research explores how social, economic, andcultural factors influence the ways in which people use, perceive,and govern natural resources, with a particular emphasis on usingapplied social science to inform coral reef management. His back-ground is in human geography and he often works closely withecologists to uncover complex linkages between social and ecologi-cal systems. He has worked on human dimensions of resourcemanagement in Jamaica, Mexico, Papua New Guinea, Kenya, Mada-gascar, Tanzania, Mauritius, Seychelles, Indonesia, Mozambique,and the USA. Dr. Cinner holds a prestigious Australian ResearchFellowship from the Australian Research Council.